Modern Approaches to Stock Price Forecasting: Analysis and Application of Machine Learning Algorithms

This paper discusses the development and application of linear regression (LR) to predict stock prices in the stock market. The main objective is to analyse the efficiency and accuracy of linear regression in relation to financial markets, and specifically to predict stock prices one day ahead. The paper begins with a detailed review of the related theoretical foundations and existing methods in the field of stock market forecasting. Next, it discusses the formulation of the stock price forecasting problem and the choice of linear regression as a tool to solve it. Particular emphasis is placed on understanding the properties of linear regression and its potential contribution to predictive models. The second part of the paper describes the implementation and testing of the LR algorithm, including the design and execution of experiments to evaluate its performance. The results are analysed using graphs and individual numerical predictions that were collected on a sample of multiple firm stocks. These results are described and discussed in detail, assessing their sensitivity to short-term volatility and deviation from fair value. The paper provides a new perspective on the use of linear regression in forecasting stock markets and contributes to a better understanding of its effectiveness and limitations.